A large body of evidence suggests that the medial temporal lobes (MTL) play a critical role in long-term memory (LTM) but little or no role in short-term memory (STM). However, recent data have challenged this standard view of MTL function by demonstrating (1) activity in the MTL when information is held in memory over short delays and (2) STM impairments in amnesic patients with MTL damage. A key function of the MTL in LTM is to bind distinct elements of experience into a unified memory so that items can later be remembered with their contextual associations (episodic memory). One possibility is that the relational binding function of the MTL extends beyond the LTM domain to support STM as well. The experiments in this application use a combined neuroimaging and neuropsychological approach to test this hypothesis in a group of amnesic patients with MTL damage and a comparison group of neurologically intact controls. The neuropsychological experiments investigate (1) whether MTL damage selectively interferes with relational STM at multiple levels of representation and (2) whether MTL damage impairs the ability to integrate across events in STM to support generalization. If amnesic patients demonstrate impairments in relational STM and generalization, this would argue against traditional neuroanatomical distinctions between STM and LTM and would provide important new insight into the role of the MTL in memory. [To compliment this neuropsychological approach, the neuroimaging procedures provide a comprehensive assessment of the neuropathology in amnesia that goes beyond standard lesion volumetric analyses. Measures of both white matter structural integrity (diffusion tensor imaging) and neuronal function (functional magnetic resonance imaging) will be collected to better characterize the local and network disruptions that may accompany damage to the MTL in amnesia. By jointly addressing questions of both cognitive and neural disruptions in amnesia, the experiments in this application will provide greater insights into the relationship between these two domains.] Linking specific measures of neuropathology to behavior will clarify the contributions of the MTL to STM and advance both mechanistic accounts of MTL function as well as theoretical accounts of STM. Because memory impairments are common in patients with neurological disorders and disease, a greater understanding of the nature of these impairments and their neural substrates will inform new strategies for reducing the burden of such illnesses. PUBLIC HEALTH RELEVANCE: This project will increase understanding of how memory representations are formed in the human brain by characterizing the nature of memory impairments in amnesia and relating these impairments to neural structure and function. In turn, such insight may serve as a foundation for improved diagnosis and more targeted treatment of memory disorders.